How To Make Figures Talk: Comparative Argument in TV Elections Night Specials
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Doury, M., 2014 : “How to make figures talk. Comparative argument in TV election night specials”, in H. Jales Ribeiro (ed.), Systematic approaches to argument by analogy, Dordrecht: Springer, p.151-169. [coordonnées actuelles : mdoury@wanadoo.fr ; feel free to ask for a reprint] How To Make Figures Talk: Comparative Argument in TV Elections Night Specials Marianne Doury CNRS, Laboratoire Commmunication et Politique Email: doury@damesme.cnrs.fr Abstract: Typically, election night specials focus on announcing the results and commenting them. These comments reveal two argumentative stages: the first consists of assessing the scores (“it’s a good / poor result”); the second is explanatory (“this poor score reflects the voters’ disapointment of the outgoing president”/ “this high score shows the voters’ longing for change”). The present paper will focus on the assessment process. The observation of election night specials during the last decades in France suggests that an electoral result is not good or bad in itself, but is discursively constructed as such. For instance, the discursive evaluation of a result is often integrated within argumentative sequences that aim at justifying it. We will examine the function of comparison in such argumentative sequences. Based on the transcript of two TV specials after the first round of French presidential elections (April, 22d, 2012) from two TV channels (TF1 and France 2), we will show how the assessing of the scores relies on various comparisons: between the results obtained by the different candidates within the same election; between the results obtained by one political party in successive presidential elections; between the results obtained by the leaders of different countries confronted with similar economic crisis; between the results predicted by polling organizations and the actual results. We aim at exploring the argumentative use of comparison as well as the associated conditions of acceptability in context. Most of the researches on comparative arguments focus on a leading subtype within this broad category, namely, figurative arguments from analogy. Figurative arguments from analogy “associate situations or cases resulting from two heterogeneous domains of knowledge” (Govier 2001: 350), and are generally based on shared non gradual properties. The purpose of this paper however is to examine arguments which also belong to the general category of comparative arguments (in that they are “based on a comparison”; Eemeren et al, 2002: 173); but they are literal comparisons, paralleling elements drawn from the same domain (Eemeren et al. 2002: 100; Brown 1995: 90), and the comparison they are based on
has a quantitative dimension: it involves properties that are numerically quantifiable. Thus,
the entities they apply to may be ranked against each other.1
My main concern will be empirical: my aim is to describe the form and to clear up the
function of a specific argument of that kind (the argumentative sequences based on a
comparison between different scores during TV election night specials), as well as the critical
discourse it may elicit; I will conclude on the difficulties raised by its analysis2.
Typically, election night specials focus on announcing the results of an election and on
commenting them. The comments usually reveal two stages.The first stage consists in
assessing the scores (“it’s a good / poor result”); the second stage is explanatory (it involves
statements such as “this poor score reflects the voters’ disappointment of the outgoing
president » or « this high score shows the voters’ longing for change”).
The present paper will focus on the assessment process. One usually feels that, at least
to some extent, the assessment of a score is determined by an objective reality: the figure that
expresses it. Nevertheless, it is hard to settle the numeric threshold beyond which a score will
clearly be considered a good one or a bad one. French citizens may remember Georges
Marchais’s unintentional sally in 1981. During the 1981 legislative elections, the number of
communist deputes had collapsed from 86 to 43. It didn’t prevent Georges Marchais, the
Communist Party’s leader, from claiming: “for me and my Party, it’s a great victory”.
Marchais’s case may be considered an atypical illustration of political bad faith: it calls
for contest from political opponents – or for laughs from the audience, as it was the case on
the TV set of TF1 after Marchais’s claim. Nevertheless, the lesson it gives holds for any
electoral score assessment: the observation of election night specials during the last decades in
France suggests that an electoral result is not good or bad in itself, but is discursively
constructed as such.
1. Presentational devices
The present research is based on an analysis of the transcript of two TV specials after the first
round of French presidential elections (April, 22nd, 2012) from two national TV channels
(TF1 and France 2).
A first exploration of the data, attentive to the discursive choices of the participants,
shows that the use of what could be called, in pragma-dialectic terms, presentational devices
(Eemeren & Houtlosser 2006), orients the scores towards a specific assessment. For instance,
the choice of an ordinal presentation rather than that of a cardinal presentation (in other
words, the choice of mentioning the rank of the candidate rather than indicating the
percentage of the votes he obtained) has clear strategic implications. As far as François
Hollande is concerned, such a choice orients towards a positive assessment of his score
(example 1), while it orients towards a negative assessment of Sarkozy’s result (example 2):
Example 1
1For a systematic inventory of the parameters used to establish subtypes within comparative arguments,
see Doury 2009.
2This paper is part of a research, initiated in collaboration with Assimakis Tseronis, which deals with
various argumentative devices aiming at making figures talk (and more specifically, at making electoral scores
talk) during election night specials on TV (Doury & Tseronis, forthcoming).Ségolène RoyalFH : c’est le candidat socialiste qui arrive euh: (...) en tête (versus “le
candidat socialiste obtient 28% des suffrages”).3
SRFH: the socialist candidate comes first (…) (versus “the socialist candidate wins 28%
of the votes”).
Example 2
Pierre MoscoviciFH : [NS] est devancé au premier tour (versus [NS] “obtient 27% des
voix”).
PMFH : [NS] is left behind in the first round (versus “[NS] wins 27% of the votes”).
A second presentational device is the choice of absolute values rather than percentages. It
orients towards a positive assessment of the score. This device is mostly used by the
candidates who obtained rather low percentages of the votes. For instance, Eva Joly, the
ecologist candidate, whose score is about two per cent4, starts her official televised statement
as follows:
Example 3
Eva Joly: Je tiens à remercier avec chaleur (...) le million de Françaises et de Français
(...) qui m’ont apporté leurs suffrages (…) et ce faisant (...) ont défendu (...) par leur
vote (...) l’écologie (...) l’Europe (…) et la République (...) exemplaire.
EJ: I want to thank warmly (…) the million French people (…) who gave me their vote
(…) and thus supported through their vote Ecologie (…) Europe (…) and the exemplary
(..) Republic.
Another example of presentational devices influencing the argumentative orientation of the
scores is the use of what Ducrot calls “argumentative operators” such as “only”, “less than” or
“close to”, which orient a claim towards a specific set of conclusions.
Thus in
Example 4
3Superscript FH, NS, MLP or JLM indicates that the politician whose name has just been mentioned
supports François Hollande, Nicolas Sarkozy, Marine Le Pen or Jean-Luc Mélenchon respectively.
4 Later on during the election night special, Gilbert Collard, an extreme-right politician, calls her
disdainfully “Mrs Two-per-cent”.Ségolène RoyalFH : (…) Monsieur Sarkozy qui ne recueille que vingt-cinq pour cent des
voix (…).
SRFH : (…) Mister Sarkozy who only polled twenty-five per cent of the votes (…).
“only” orients towards a negative assessment of the score. In
Example 5
Journalist: Monsieur Mélenchon vous faites moins de douze pour cent [11,7% à ce
moment de la soirée]. Est-ce que vous êtes déçu?
J: Mister Mélenchon you win less than twelve per cent of the votes. Are you
disappointed?
“less than” has the same argumentative orientation, whereas in
Example 6
Journalist (J): Marine Le Pen fait un score proche de vingt pour cent [18,6% à ce
moment-là de la soirée] est-ce qu’on peut considérer qu’vous avez perdu votre bras
d’fer avec elle ?
J: Marine Le Pen obtains a score close to twenty per cent can we say that you lost your
tug-of-war game against her?
“close to” orients to a positive set of assessments.
Such presentational choices give the score a discursive form which makes it possible for it to
serve as a premise for a specific conclusive assessment. They orient the numerical results and
prepare them to support preferentially certain conclusions over others.
Comparisons
Beyond the presentational devices, the discursive evaluation of a result is often integrated
within argumentative sequences that aim at justifying it. In this section, the function of
comparison in such argumentative sequences will be investigated. As will be shown below,
the assessment of the scores relies on various comparisons: between the results obtained by
the different candidates within the same election; between the results obtained by one political
party in successive presidential elections; between the results obtained by the leaders of
different countries confronted with similar economic crisis; between the results predicted bypolling organizations and the actual results. The argumentative use of such comparisons will
be highlighted, as well as the associated conditions of acceptability in context.
The argumentative comparison we will focus on may be described, at the most
superficial level, as a shift, from the relative assessment of two scores, to the absolute
assessment of one of them:
S1 > S2 à S1 is a good score
S1 < S2 à S2 is a poor score
Such a scheme results in various specific comparisons, according to the nature of the phoros.
Comparisons between the scores of the candidates in the same election
The most obvious comparison in the context of a presidential election first round is between
all the candidates that took part in this round. Such a comparison allows for the ranking of the
candidates with regards to the scores they obtained.
Objectively, the main political relevance of such a ranking is to identify the first two
candidates as opposed to all the other ones: they, and only they, will take part in the second
round of the presidential elections two weeks later. The observation of former presidential
elections in France shows that the ranking of the first two candidates after the first round does
not presage of the final result of the election. Hence this ranking has only a symbolic
importance. Nevertheless, as suggested by examples 1 and 2, François Hollande’s supporters,
rather than specifying the score obtained by their candidate, prefer only mentioning that he is
the leader of the election over his challenger, whom he “distinctly gets ahead of”:
Example 7
Jean-Marc AyraultFH: C’est clair, c’est un désaveu incroyable du président sortant (...)
il [Nicolas Sarkozy] arrive en deuxième position (...) en plaçant nettement François
Hollande en (...) en première place (…) parce qu’il incarne (…) le changement (...) et le
redressement de la France.
JMAFH: Clearly it’s an amazing disavowal for the Past President (…) he [Nicolas
Sarkozy] comes in second (...) placing clearly François Hollande in first position (...)
because he incarnates change and recovery for France.
There are two possible reactions to such an argument.
The first one consists in minimizing the difference between the score obtained by
François Hollande and that obtained by Nicolas Sarkozy, as in examples 8 and 9.
Example 8
Nadine MoranoNS : Ils sont arrivés (…) dans un mouchoir de poche.NMNS : It was a close finish [French expression is literally: they arrived in a pocket
handkerchief].
Example 9
Henri GuainoNS: Il [NS] arrive pas loin hein ?
HGNS: He [Nicolas Sarkozy] does not come far behind, does he ?
Such reactions suggest that for the arguer, the order of arrival of the candidates does not
suffice to conclude to a positive or a negative absolute assessment of their scores: an
important difference between the scores is needed (and of course, Sarkozy’s supporters
consider that the difference between their candidate and Hollande is not important enough to
support Jean-Marc Ayrault’s conclusion that “it’s an amazing disavowal” for Sarkozy).
The second possible reaction to such a comparative argument, which is clearly
unfavorable to Sarkozy, is to propose another comparison favorable to him, as in example 10.
In reaction to Hollande’s supporters’ emphasis on the fact that Hollande has overtaken
Sarkozy, Valérie Pécresse claims that:
Example 10
Valérie PécresseNS: Nicolas Sarkozy (...) a réussi à devancer le Front National.
VPNS: Nicolas Sarkozy (…) has succeeded in overtaking the National Front.
Valérie Pécresse thus illustrates a property common to all comparative argument schemes:
they can always be opposed another comparison orienting towards a different, if not opposite,
conclusion.
However, the strategic possibility of choosing another reference candidate in order to
make the comparison favorable to one’s candidate is limited by the requirement that the
comparison should be politically significant. It is the case when Marine Le Pen’s supporters
congratulate themselves that her score is two times Jean-Luc Mélenchon’s score: as Jean-Luc
Mélenchon made defeating Marine le Pen a stake of the elections, the comparison makes
sense. By way of contrast, Marine Le Pen’s score is a hundred times Jacques Cheminade’s
score, who obtains 0.2% of the votes; though it would undoubtedly have been spectacular, the
comparison is never raised by Marine Le Pen’s supporters because it definitely lacks any
political relevance. Hence political relevance should be considered another requirement for
such a comparison to hold.
Comparison between a candidate’s score and his (or his party’s) previous scores during
former presidential elections
Another kind of comparison consists in comparing the score of a candidate to the previous
scores he or his party obtained during former presidential elections. Such a comparison still
aims at justifying an assessment of the score, which constitutes the conclusion of the
comparative argument. This comparative device may be used by a politician in order to
enhance the result of his favorite candidate (example 11), or in order to depreciate that of his
opponent.Example 11
Ségolène RoyalFH: Ce score (...) est exceptionnel [conclusion] puisqu’il dépasse en effet
le score de deux mille sept (...). [Argument]
SRFH : This score (…) is exceptional [conclusion] in that it exceeds the two thousand
and seven score (..). [Argument]
It also is a means for journalists and commentators to exhibit their political analyst’s
expertise, and to demonstrate that they are depositories of the political memory of the country;
therefore they often base their commentaries on such diachronic comparisons.
Besides the mechanism shared by all the comparisons described here, the diachronic
dimension of this one allows the arguer to endow the score obtained by a candidate with a
positive or a negative dynamics, and to suggest that he or she is on an upward or downward
path – which is particularly important at the night of a first round election in prevision of the
second round coming soon.
Such comparisons allow politicians whose results are deemed disappointing by
commentators to evade a negative absolute assessment owing to the evocation of a positive
relative evaluation. Thus Pierre Laurent, from Front de gauche (radical leftist coalition),
invited to comment on Jean-Luc Mélenchon’s relatively poor score, answers as follows:
Example 12
Pierre LaurentJLM: C’est à la fois un très bon résultat [conclusion] parce que je rappelle
que (…) pour la gauche: que nous représentons (...) c’est le meilleur résultat ET de loin
depuis trente ans [argument] (…) c’est un très beau résultat. [Conclusion].
PLJLM: At the same time it is a very good result [conclusion] because I remind you that
(…) for the left we represent (...) it’s been by far the best score for thirty years
[argument] (…) It’s a very beautiful result. [Conclusion].
The same kind of comparison, when applied to Nicolas Sarkozy, suggests a reverse evolution
– that is, an evolution that follows a downward path. Hence Arnaud Montebourg, one of
Hollande’s supporters, does not hold back from using it:
Example 13
Arnaud MontebourgFH: Euh d’abord un président sortant sanctionné et très lourdement
[conclusion] euh cinq à six points en moins qu’il y a cinq ans (…). [Argument (FR2)]
AMFH: Uh first a Past President severely penalized [conclusion] uh five or six points
below the score he obtained five years ago (...). [Argument (FR2)]
Once again the comparison is used as an argument to support an interpretation of the first
round results as a penalty against Nicolas Sarkozy.A recurring strategy in order to resist such an argumentation consists in contesting the
comparison by claiming that you can compare only what is comparable. In this specific case,
the comparison between the results obtained by Sarkozy and those obtained by any former
outgoing president would not hold because of the economic context: the present exceptional
situation of crisis would make it impossible to compare the present election with any previous
one.
The examination of the two election night specials also shows that a comparison may be
used by journalists to excite rivalries between members of the same political camp when a
party presented different candidates to successive elections. If the comparison is favorable to
one candidate, its evocation may be uncomfortable for the others.
It is particularly the case when the comparison involves two candidates connected by personal
links, like Marine Le Pen, the present candidate of the extreme right Party, and Jean-Marie Le
Pen, her father and former candidate of the National Front:
Example 14
Journalist (à JMLP): Elle [MLP] fait un score encore plus important qu’vous en deux
mille deux/ est-ce que vous avez l’impression qu’elle vous éclipse.
J (to Jean-Marie Le Pen): The score she [Marine Le Pen] obtains is even more
important than yours in two thousand and two don’t you feel like she’s eclipsing you.
Jean-Marie Le Pen’s reaction consists in challenging the relevance of the comparison:
according to him, he and his daughter should not be considered as rivals (as a comparing
process, according to him, would imply) but as team-mates in a relay race: he ran first, she
took over, and what counts is the result of the team, not that of each runner:
Example 15
Jean-Marie Le PenMLP: Pas du tout (...) c’est un relayeur Marine (...) je lui ai passé le
bâton (...) elle court, elle court plus vite tant mieux [sourire].
JMLPMLP: Not at all (...) Marine is a relay runner (.) I gave her the baton (...) she runs,
she runs faster all the better [smile].
He thus claims for the shared responsibility of Marine Le Pen’s high score, and accordingly
suggests that a comparative process does not fit the way the situation should be framed.
A very similar script may be observed as regards Ségolène Royal, the former socialist
candidate to the 2007 presidential elections, and François Hollande, her former partner and
candidate in 2012:
Example 16
Journalist: Pardon de vous poser cette question, mais c’est une soirée un petit peu
particulière j’imagine pour vous, puisque (...) votre ancien compagnon arrive en tête ilfait un meilleur score que vous aux à la dernière élection présidentielle (…) est-ce que
vous en avez parlé avec lui est-ce que vous l’avez eu au téléphone ce soir ?
J: Sorry for asking you this question, but this evening is quite special for you I guess
because (...) your former partner comes in first he gets a better score than you did to the
last presidential election (…) did you talk with him about that did you talk to him by
phone tonight?
The journalist’s comparison clearly aims at tickling Ségolène Royal’s pride (if the comparison
orients to a positive assessment of Hollande’s score, it also orients to a negative assessment of
Royal’s former score). Ségolène Royal reacts as follows:
Example 17
Ségolène RoyalFH: Oui, bien sûr (...) et d’ailleurs c’est précisément pourquoi je suis bien
placée pour dire que ce score (…) est exceptionnel puisqu’il dépasse en effet le score de
deux mille sept (…) et en même temps (…) euh il y a un score du Front National qui
double (...) par rapport à deux mille sept (.) ce qui fait que le combat (...) euh de l’entre-
deux tours va être un combat (...) decisif que nous devons gagner pour les Français.
SRFH: Yes, of course (…) and it is precisely the reason why I am in a good position to
state that this score (…) is exceptional for it overtakes indeed the 2007 score (…) and at
the same time (…) uh we have the score of the National Front that doubles (...) with
regard to 2007 (…) it follows that the fight between the two rounds will be a decisive
fight and we have to win it for the French people.
The comparison by the journalist between Royal’s and Hollande’s scores is introduced as an
argument for the vague conclusion “this evening must be quite special for you”, conclusion
which might easily be specified as “it must be a hard blow for you tonight”. Royal turns the
comparison into an argumentatively opposite direction, making an argument out of it that
increases the legitimacy of her positive assessment of Hollande’s performance. Then she turns
to another comparison (the National Front doubling its former score) which is much more
spectacular and thus downgrades the progress performed by Hollande. Besides, it diverts the
positive assessment induced by Hollande’s progression on to an anxious concern about the
dramatic rise of the National Front.
It is worth noting that in the context of the first round election night specials, the
stability of the scores obtained by one candidate or one party from one election to another is
seen as a sign of stagnation and is integrated within disqualifying strategies. Thus, Jean-
François Copé claims that the left obtains the same results as in two thousand and seven (c’est
“la même chose qu’en 2007”) in order to downgrade François Hollande’s first position; and
stability from one election to another is never presented as a positive outcome by the
candidates and their supporters. Alain Juppé adopts the same disqualifying strategy:
Example 18Alain JuppéNS: Deuxièmement (...) il n’y a pas de poussée de gauche (…) hein, quand
on totalise toutes les voix d’gauche on est à quarante-deux et d’mi pour cent (…)
comme (...) en deux mille sept.
AJNS : Second (...) there is no upsurge of the left (...) when you add all the votes of the
left you get forty-two percent point five (…) as (…) as in two thousand and seven.
Contrary to what this example illustrates, most of the time during the two electoral night
specials studied here, Sarkozy’s supporters content themselves with claiming that “there is no
upsurge of the left” without mentioning any precise score. This may be due to the fact that the
comparison of the scores obtained by the left in 2007 and 2012 does indeed indicate an
unquestionably ascending dynamics (from 36,7% in 2007 to 44% of the votes for the left in
2012): hence there is a clear interest for Sarkozy’s supporters to remain vague when assessing
that “there is no upsurge of the left”, and to leave unspecified which left they are talking
about, or which former election they refer to.
Comparison between a candidate’s score and the results obtained by the leaders of
different countries confronted with similar economic crisis
As mentioned before, an argumentative comparison which does not serve one’s position can
always be opposed another anti-oriented comparison. Thus Arnaud Montebourg’s former
diachronic comparison, which shows Sarkozy disadvantage (ex. 13), is opposed Jean-François
Copé’s alternative synchronic comparison, which allows him to assess positively the past
President’s score:
Example 19
Jean-François CopéNS : Vous savez (…) moi, je vais vous dire une chose (...), euh, c’est
(…) dans cette période de crise (...) j’ai vu (…) dans tous les grands pays où y a eu des
élections (...) euh, ils, les responsables sortants, ont été balayés rappelez-vous en
Espagne Monsieur Zapatero (...) Monsieur Gordon Brown en Angleterre (...) balayés
(...) là (…) y a quoi (...) un point et demi d’différence entre les deux.
JFCNS: You know (...) I’ll tell you something (...) uh, it’s (...) in this period of crisis (...).
I saw (…) in all big countries where elections have taken place (…) uh the outgoing
leaders have been swept away remember Mr Zapatero in Spain (…) Mr Gordon Brown
in England (...) swept away (...) here (...) there’s let’s say (...) one point and a half
between the two candidates.
The score obtained by Nicolas Sarkozy is poor if compared to the previous presidential
elections; but it is superior to the results obtained by the past leaders in the European
countries undergoing similar economic crisis; and in this light, Sarkozy’s score is a good
score.Comparison between the results predicted by the very candidate and the actual results
The comparison is ensured to escape any refutation of this kind when it compares the actual
score of a candidate to the score that was explicitly aimed at by this very candidate (or by his
camp).
This argument is used by Hollande’s supporters to make Sarkozy’s second place
meaningful: such a second place is synonymous with defeat, since Sarkozy himself explicitly
ambitioned to arrive first:
Example 20
Ségolène RoyalFH: C’est la première fois qu’un: candidat non sortant (...) euh, arrive en
tête (…) donc c’est c’qu’il faut retenir, puisque monsieur Sarkozy nous avait annoncé
qu’il était en tête là (...) c’est le candidat socialiste qui arrive, euh (...) en tête.
SRFH : It’s the first time that a non-outgoing candidate (…), uh, arrives first (…) so it’s
what should be remembered since Mr Sarkozy had announced that he would come first
here (...) it’s the socialist candidate who comes first.
Sarkozy’s own ambitions are frequently evoked all along the night in order to justify a
negative assessment of his actual score, as in examples 21 to 23:
Example 21
Ségolène RoyalFH (à AJ): Mais vous l’avez placé comme objectif d’être premier au
premier tour.
SRFH (to AJ): But you had stated it as an objective to come first at first round.
Example 22
Journalist: Nicolas Sarkozy (...) est beaucoup plus bas qu’il ne l’espérait.
J: Nicolas Sarkozy (…) is much below what he hoped.
Example 23
Marie-Georges BuffetJLM/FH : Encore hier vous disiez que Sarkozy serait premier
Monsieur Copé.
MGBJLM/FH: Yesterday you still claimed that Sarkozy would come first Mr. Copé.
The same strategy applies to other candidates who imprudently set themselves objectives to
reach. This is the case for Mélenchon, who aimed at overtaking Marine Le Pen and who
failed, as Louis Alliot (who supports Marine Le Pen) stresses:
Example 24Louis AlliotMLP : [à Clémentine AutainJLM] C’est un véritable échec de Jean-Luc
Mélenchon (…) vous le reconnaissez ou pas (...) Le seul défi du Front d’gauche, c’était
de faire plus que Marine Le Pen ben c’est loupé complètement.
LAMLP : [to CAJLM] It is a real failure for Jean-Luc Mélenchon (…) do you admit that or
not (…) the only challenge for the Front de gauche was to get a better score than Marine
Le Pen so it’s a complete failure.
Such comparisons are based on an ad hominem argument in Perelman’s sense (Perelman &
Olbrechts-Tyteca 1988: 148), that is, an ex concessis reasoning: the conclusion is all the more
solid since the premises are drawn from the opponent. Such a result is bad with regards to the
standard the opponent himself proposed.
In such cases, the only solution left for the attacked politicians to counter such
comparisons is to suggest that the candidate had to fight headwind (because of the economic
crisis, because of a shameful campaign, etc.): it’s a way of invoking mitigating circumstances,
but it does not contest the acknowledgment of failure.
Comparison between the results predicted by polling organizations and the actual
results
A fifth kind of comparison used to make figures talk during election night specials is
somewhat puzzling. It consists in comparing the actual results of a candidate with the results
that were predicted by polling organizations and that were spread by the media.
In France in recent years, criticizing the accuracy and the political effects of the surveys
achieved by polling organizations has become a national sport. This sport is intensively
practiced during election night specials such as the ones studied here. The present paper
proposes to connect poll criticism during election night specials with general argumentative
strategies and local political stakes.
Examples 25 to 27 below are nothing but additional illustrations of the argumentative
mechanism at the core of this paper, which consists in shifting from the relative assessment of
two scores, on the basis of a comparison, to an absolute assessment of one of them.
Here Alain Juppé and Xavier Bertrand, who support Nicolas Sarkozy, attempt to
downgrade the results of one of his opponents (Jean-Luc Mélenchon, the leader of the Front
de gauche). Comparing Mélenchon’s score with his party’s former score would be favorable
to him (it would show a spectacular uprise); thus such a comparison must be avoided. On the
contrary, polling organizations often predicted that the Front de gauche’s score would be
higher than it actually is. So such a comparison is of interest to Sarkozy’s supporters, who
often use it in order to justify the negative assessment of Mélenchon’s score:
Example 25
AJNS: Moi je dirais que (…) l’élection ne ressemble pas à ce qu’on nous avait annoncé
(…) le score mirifique annoncé à monsieur Mélenchon/ (…) n’a pas été au rendez-vous.AJNS: personally I would say that […] the election does not look like what was
announced (…) the fantastic score predicted for Mister Mélenchon (...) has not
materialized.
Example 26
XBNS:[Sur Nicolas Sarkozy] le score (…) est un score qui est solide (...) et d’ailleurs on
l’voit bien (...) beaucoup pensaient que c’était déjà joué ; rappelez-vous voilà quelques
semaines (...) il serait même pas qualifié au second tour paraît-il (...) il) serait même
pas (...) dans le duo de tête (...) et là aujourd’hui (...) j’ai juste de la mémoire.
XBNS : [On Nicolas Sarkozy] the score (...) is a solid one (…) and besides you can see it
(…) many people thought everything was already settled remember a few weeks ago
(...) he would not even qualify for the second round they said (...) he wouldn’t even be
one of the top two (…) and now today (...) I just have a good memory.
Example 27
AJNS: Enfin il faut quand même se souvenir (...) que il y a trois mois (…) Nicolas
Sarkozy était donné absent (...) du deuxième tour (…) on le mettait à vingt pour cent
(...) et il a fait une campagne qui lui a permis de progresser d’façon très significative
(…) c’est assez exceptionnel pour un chef d’Etat dans une période de crise (...).
AJNS: Now one should remember (…) that three months ago (...) Nicolas Sarkozy was
predicted absent (…) from the second round (…) he was assessed at about twenty per
cent (...) and owing to his campaign he progressed in a significant way (.) it is quite
exceptional for a head of state in a period of crisis (...).
In these examples it is not always clear if the polling organizations are said to have been
mistaken (it seems to be the preferred interpretation of ex. 25) or if the predicted score was a
correct reflect of public opinion at the time the poll was made, but the electors’ voting
intentions have changed (example 27). In the latter case, the argumentative comparison
resembles a former case of comparison: that between the successive scores of the same
candidate or party.
At any rate, the comparison between the actual scores and those announced by the polls
obeys the same argumentative logic as the one described so far. Jean-Luc Mélenchon’s score
is inferior to what was predicted: it can thus be deemed a bad score. Sarkozy’s result is
superior to what was announced: in this regard, it may be considered a good score.
The same process is applied to François Hollande by Rachida Dati in
Example 28
RDNS: Tout d’abord moi je trouve que aujourd’hui il faut rendre hommage aux Français
qui euh qui se sont mobilisés (...) avec gravité/responsabilité (…) ils ont déjoué tous les
pronostics (...) des sondeurs mais aussi des médias (...) d’abord on passe pour Monsieur
Hollande de trente-huit pour cent à vingt-huit pour cent (…) euh, monsieur Teinturier
(...) dans les sondages (…) depuis le mois de novembre (…) Hollande n’était pas à
vingt-huit pour cent il y a quelque temps.RDNS: First of all personally I think that we should pay homage to the French people
who joined forces (.) with gravity/ responsibility (...) they thwarted all forecasts (...) of
the polling organizations but also of the media (…) first it goes for Mister Hollande
from thirty-eight per cent to twenty-eight per cent (...) uh, Mister Teinturier [a pollster
from IPSOS] in opinion polls since November (…) Hollande was not given twenty per
cent some time ago.
In this example as in the previous ones, the opinion polls are not explicitly connected to a
specific polling organization or to a specific media, and are not assigned specific dates; such
characteristics reinforce the hypothesis that their evocation is merely meant to produce a
strategic upgrading or downgrading effect on the score of a candidate the speaker wants to
support or to attack, but it is not sincerely laden with any deep political significance.
Such a vague account of the opinion poll’s reference makes it difficult for Rachida
Dati’s targets to refute her claims. Example 29 shows how Brice Teinturier, the pollster from
Ipsos who was mentioned by her, contests her attack, and how she reacts in turn:
Example 29
Brice Teinturier: Soyons très précis et très factuels. D’abord aucun sondage n’a donné,
euh (…) François Hollande à trente-six pour cent. Nous étions à vingt-neuf pour cent au
début janvier (...) trente-deux pour cent au moment du Bourget mais tout cela est (...)
parfaitement vérifiable (…) et vingt-neuf pour cent dans l’dernier enquête (...) il ferait
vingt-huit et demi. On est très près.
RD: Mais c’est au moment des primaires en fait.
BT: Let’s be very accurate and very factual. First no poll gave uh (...) François Hollande
at thirty-six per cent. We were about twenty-nine per cent in early January (…) thirty-
two per cent at the time of the Bourget Meeting, but all this can easily be checked (...),
and twenty-nine per cent in the last survey (…) now he obtains twenty-eight point five .
We’re very close.
RD: Actually I was referring to the primary election period.
In France as in many countries, opinion surveys aiming at assessing the intentions of vote
before important elections are so numerous that one can probably always find a support for
the comparison to work by simply altering the date or the source of the poll which is being
invoked.
When introducing this fifth kind of comparison, I described it as “puzzling”. All
previous comparisons may be questioned and challenged to some extent; but this one seems
hardly reasonable. How can one consider that the fact that an electoral result is better than
what the polls predicted makes this result better “in the absolute” (specifically when the
difference between the predicted results and the actual ones are due to an error of the polling
organizations, as suggested in example 25, rather than to a change in public opinion)?Critical discourse on comparison
This last remark opens the discussion to the issue of the critics of comparison.
I will first handle it from the point of view of the arguers, and I will recapitulate the various
reactions aroused by the comparisons we’ve been examining. I will then endorse the analyst’s
perspective and try and match the actual examples with specific comparative argumentative
schemes – and in particular, a fortiori argument.
I will first propose to consider the arguer’s reactions as reflecting a spontaneous normative
stance towards this kind of comparative argument. These reactions may be listed as follows:
- A first reaction to comparisons consists in contesting the truth of the premises, and
particularly, the score used as a phoros in the comparison (ex. 29). Of course, such a
critical strategy is not specific to comparative arguments, in that the rebuttal of many
other argument schemes may follow this line; nevertheless, it makes sense as far as
inductive analogies (that are empirically based; Govier 1989: 141) are concerned, but it
is bluntly irrelevant as regards a priori analogies (in which the phoros may be fictitious;
Govier 1989: 142).
- Another reaction is to suggest that the difference between the compared scores is not
“significant enough” to warrant the soundness of the conclusion: Sarkozy’s supporters’
claim that the two top candidates arrived shoulder to shoulder in order to contest any
positive conclusion about Hollande’s leadership over Sarkozy (ex. 8-9).
- Challenging the comparison may amount to suggesting that important differences
prevent the comparison from being conclusive (an outgoing president is being compared
to an outgoing prime Minister, elections in a situation of economic crisis are being
compared to elections in a much more favorable context):
Example 30
Valérie PécresseNS : La dernière fois qu’vous avez gouverné la France c’était pas (...)
avec une crise mondiale (...). Vous n’étiez pas qualifié pour le second tour. Nicolas
Sarkozy (...) a réussi sa qualification pour le second tour (…). Il a réussi à devancer le
Front National.
VPNS : Last time you governed France it was not (...) in a situation of worldwide crisis
(…). You were not qualified for the second round Nicolas Sarkozy (...) did succeed in
qualifying for the second round of the elections (…) he succeeded in getting ahead of
the National Front.
-A very common reaction to comparisons consists in proposing an anti-oriented comparison
(ex. 10, 19), presented as more conclusive.
- At last, another way of handling the issue is to suggest that comparative arguments reflect an
erroneous framing of the political stakes of the election (ex. 15).
Some of these reactions echo the critical questions associated with comparative arguments
(see for instance Eemeren & Grootendorst 1992: 101-102). This is why, even if I agree with
Perelman’s claim that what he calls “arguments from comparison” (or what Plantin (2011)calls an “analogy of comparison” [une analogie de comparaison]) show important differences
with arguments from analogy, I don’t follow him when he considers them as “quasi-logic
arguments”, that is, as belonging to a radically different category of arguments than
arguments from analogy (which according to Perelman belong to arguments that establish the
structure of reality; Perelman & Olbrechts-Tyteca 1988: 326). In other words, I claim that the
arguments from comparison studied here are a subcategory of comparative arguments, as
arguments from analogy strict sensu are.
Reconstructing such arguments
Reconstructing empirical arguments along the lines of classically identified argument
schemes makes their description more accurate, and, in a normative perspective, makes it
possible for a systematic and rigorous assessment to proceed.
Hence the superficial initial description of the mechanism studied here (an argument
which allows to depart from a relative assessment of two scores to the absolute assessment of
one of them) should be connected to a structural description according to specific argument
schemes identified within Argumentation Studies.
Eemeren, Grootendorst & Snoeck-Henkemans (2002) propose a very general description of
comparative arguments, or arguments based on a relation of analogy in the broader sense of
the term:
Y is true of X
Because: Y is true of Z,
And : Z is comparable to X.
When applied to the examples examined here, such a description produces the following
reconstruction:
S2 is a good/bad score
Because: S1 is a good/bad score
And: S2 is comparable to S1
An obvious flaw of such a reconstruction is that it misses the specific orientation of the
comparisons studied here, which more often than not focus on the difference between S1 and
S2 rather than on their closeness. The arguments studied here do not presuppose that S2 is
similar to S1: they claim that S2 is superior, or inferior, to S1, as a premise for a conclusive
assessment of S2. In that, they echo Brown’s penetrating, though unexploited remark, that
“although usually not mentioned (…), differences would seem to be as susceptible as
similarities to analogical treatment.” (Brown 1989: 162).
What about an a fortiori reconstruction?
Hence, as the comparisons studied here involve some kind of hierarchical organization (a
score is said to be superior or inferior to another one, not only “similar”), the a fortiori
argument, which is often presented as a subtype of comparative arguments, and which implies
a gradual dimension, might well be a more relevant frame for the analyst to characterize the
structure of our examples.In connection with the a fortiori argument scheme, the description of the argumentative
mechanism involved in the examples we’ve been examining so far requires that one uses
Ducrot’s notion of argumentative scale (Ducrot 1973: 229). An argumentative scale is a set of
arguments oriented to supporting a conclusion; these arguments are ordered according to their
argumentative strength.
Fig. 1:
C : Marie is a learned person
An (Marie can (even) read Sanskrit)
A…
A2 (Marie can read ancient greek)
A1 (Marie can read latin)
An argument An is stronger than an argument A1 means, in Ducrot’s perspective, that if one
accepts to conclude from A1 to C, then (a fortiori) he must also accept to conclude from An to
C.
Hence the connection between Ducrot’s definition of an argumentative scale and the a
fortiori argument is direct. The argumentative scale corresponding to our data is:
Fig. 2:
C : S is a good score
Sn
S…
S2
S1
The a fortiori argument based on such an argumentative scale would run as follows:
S2 is a good score
Because: Sn> S1
And: [S1 is a good score]Unfortunately, such an argumentative structure cannot always be applied as it is to our
examples. As it stands, the description of the a fortiori argument presupposes that the lower
claims on the scale (here, S1), even if weaker than the arguments at the top of the scale, still
have to be seen as arguments for C.
Such a presupposition fits some examples of the data. For instance, when Marine Le
Pen’s score is compared to her father’s previous score in 2002, it is clearly consensually
assumed that her father’s score was deemed a high score (ex. 31), and sometimes even made
explicit (ex.32):
Example 31
FN: Bien (…), ce qu’on retiendra, euh (…), surtout ce soir au soir du premier tour c’est
le score (...) impressionnant de Marine Le Pen, euh, vingt pour cent des voix c’est’à dire
(...) bien au-delà du score réalisé par son père (...), euh, au soir du premier tour du vingt-
et-un avril deux mille deux.
FN: Well (...), what we will retain, uh (...), especially tonight after the first round is the
(...) impressive score of of Marine le Pen, uh, twenty per cent of votes that is (...) far
beyond the score that her father had obtained (…), uh, at the first round evening of the
twenty-first of April two thousand and two.
Example 32
Sympathisant: Elle explose le score historique (...) du Front National (...) et
effectivement c’est un très, très grand soir (…) pour les patriotes pour les
souverainistes.
A supporter: She sweeps away the historical score (…) of the National Front (...) and in
fact it is a very very great evening (...) for patriots for sovereignists.
In this example the positive assessment of Marine Le Pen’s score is secured by the
comparison with her father’s previous score, and by the positive assessment of the latter: if
Jean-Marie Le Pen’s 2002 score is deemed a good score, then Marine Le Pen’s score, which
is superior to it, may all the more be deemed so.
The relevance of the evaluative premise [S1 is a good score] is confirmed a contrario
by the fact that when such a premise is clearly unacceptable for all participants, the inference
from the comparison to the conclusion is held up. This is how example 33 should be
understood:
Example 33Claire Chazal (journaliste) [À Cécile Duflot] : Vous faites un peu plus de deux pour
cent; c’est légèrement mieux qu’la dernière fois, mais c’est un score qui reste faible.
Claire Chazal (journalist) [To Cécile Duflot, an ecologist activist] : You win a little
more than two per cent it is slightly better than last time but it is still a low score.
In a Ducrot’s style analysis, the use of “but” signs the fact that “it is slightly better than last
time” would normally point to a conclusion like “it’s a good score” as long as the phora score
can be considered a good one – or at least, not too poor a score. But in the case of the
Ecologist party, the former score (1.57%) was so low that the fact that the present score is
superior to it does not suffice for concluding that “it’s a good score”.
Along the same line, if S1 is deemed an extremely high score, a comparison unfavorable
to S2 (stating S2’s inferiority relatively to S1) can still authorize a positive absolute
assessment of S2 – as long as the presentational devices used to phrase the comparison fit the
argumentative objective (for instance “tonight Candidate X almost makes it as well as he did
during previous elections”, as opposed to “tonight Candidate X does not even reach the score
he obtained during previous elections”).
But most of the time, the second premise (“S1 is a good/bad score”) is left implicit. And
reconstructing it as an implicit premise would sometimes unduly “force” the arguer’s
reasoning into a pre-established argument scheme whereas nothing, in the very phrasing of
the argument, points to the relevance of such an interpretation. In these cases everything runs
as if the comparison argument were based on a multi-directional argumentative scale as in
Fig. 3:
S is a good score
Sn
S…
S2
S1
S is a bad score
In such cases the only important premise to conclude that S2 is a good score is that S2>S1 (or
to conclude that S2 is a poor score is that S2To conclude with the analytic account of the comparative structures studied here, one
should add that the a fortiori reconstruction of the examples sometimes requires that the
enunciative, polyphonic dimension of discourse be taken into account. It is the case for
example 34, which runs as follows:
Example 34
Gilbert CollardMLP: Bon (...) vous avez entendu m’sieur Mélenchon dire qu’avec onze
virgule six (...) il est la clé du futur (...) alors (...) que sera (…) Marine Le Pen avec dix-
neuf euh virgule: six à l’heure actuelle hein (...) la la le serrurier de la vie politique (…)
euh la clé de l’avenir (...) à droite (...). C’est Marine Le Pen.
GCMLP : Well (...) you heard Mr Mélenchon saying that with eleven point six (...) he is
the key of the future (...) then (…) what will be Marine Le Pen with nineteen uh point
six at the moment (…) the the locksmith of political life (...) uh the key of the future
(…) to the right (…). Is Marine Le Pen.
This example fits to some extent the usual description of a fortiori arguments. It suggests that
within the set of the arguments for the conclusion “Candidate X is the key of the future”, “to
get 19.6 of the votes” is stronger than “to get 11.6 of the votes”; if one accepts to conclude
from “candidate X got 11.6 of the votes” to “candidate X is the key of the future”, then, a
fortiori, one also must accept to conclude from “candidate X got 19.6 per cent of the votes” to
“candidate X is the key of the future”.
The problem lies in the absolute phrasing of the conclusion: “to be THE key of the
future”. This claim cannot be true simultaneously of Jean-Luc Mélenchon and of Marine Le
Pen. Hence the a fortiori argument reconstruction can be understood only resorting to a
polyphonic interpretation of it: the inference from “candidate X got 11.6 % of the vote” to
“candidate X is the key of the future” is attributed to Mélenchon’s supporters, whereas the
inference from “X got 19.6 % of the votes” is endorsed by Gilbert Collard himself. Collard’s
argument seems to be based on an enunciatively complex argumentative scale, as represented
in
Fig. 4:
In Mélenchon’s perspective / In Collard’s perspective
S is a good score
Sn Sn
19.6% 19.6%
Argumentative reversal
threshhold
Argumentative 11.6% 11.6%
reversal threshold S1 S1
S is a bad scoreUndoubtedly, the present paper proposes a systematic account of an argumentative device (the argumentative use of comparisons between scores during election night specials) which is so specific that it might well seem anecdotal; but such a judgment would be mistaken, insofar as resorting to comparisons is a very common strategy used in order to make sense of quantitative data in all sorts of circumstances. The examples we’ve been through suggest that whereas the conclusive evaluations (“it’s a good score” / “it’s a bad score”) that follow from the comparisons between different scores sometimes seem to be reasonable – at least to some extent -, in other cases the comparisons seem to have a merely mechanical upgrading or downgrading effect that is more attached to the superficial comparative structure than to any reasonable political reasoning; in this sense, the argumentative comparisons studied here illustrate the complex connection between linguistic phrasing and the logical base of argumentation. References Brown, W. R. (1989). Two traditions of analogy. Informal Logic XI: 161-172. Brown, W. R. (1995). The domain constraint on analogy and analogical argument. Informal Logic 17(1): 89-100. Doury, M. (2009). Argument schemes typologies in practice: The case of comparative arguments. In F. H. van Eemeren & B. Garssen (Eds.), Pondering on Problems of Argumentation: Twenty Essays on Theoretical Issues (141-155), Dordrecht: Springer. Doury, M., & Tseronis, A. (forthcoming 2014). Les faits et les arguments: La mise en discours des scores électoraux. In J. Goes, J.-M. Mangiante, F. Olmo & C. Pineira (Eds.), Actes du colloque “Le langage manipulateur: pourquoi et comment argumenter?”, Arras, 13- 15 septembre 2012. Arras : Artois Presse Université (APU). Ducrot, O. (1973). La preuve et le dire. Paris: Mame. Eemeren, F. H. van, & Grootendorst, R. (1992). Argumentation, Communication, and Fallacies: A Pragma-dialectical Perspective. Hillsdale, New Jersey: Lawrence Erlbaum. Eemeren, F. H. van, Grootendorst, R., & Snoeck-Henkemans, F. (2002). Argumentation. Analysis, Evaluation, Presentation. Mahwah, NJ: Lawrence Erlbaum Associates. Eemeren, F.H. van, & Houtlosser, P. (2006). Strategic maneuvering: A synthetic recapitulation. Argumentation 20: 381-392. Govier, T. (1989). Analogies and missing premises. Informal Logic 11(3):141-152. Govier, T. (2001). A Practical Study of Argument. 5th edition. Belmont: Wadsworth. Perelman, C., & Olbrechts-Tyteca, L. (1958/1988). Traité de l’argumentation. La nouvelle rhétorique. Bruxelles: Editions de l’Université de Bruxelles. Plantin, C. (2011). Analogie et métaphore argumentatives. A Contrario 16: 110-130.
You can also read
